Is Claude Code down?
How to check Claude Code's status properly
"Claude Code" is its own component on Anthropic's status page — it can be degraded while claude.ai chat is fine, and vice versa.
"Claude API" covers api.anthropic.com. If it's down, Claude Code is down with it, along with every tool that calls Claude.
Model-level incidents ("Elevated errors on Claude …") often affect one model family only — switching models can be enough.
What Claude Code's incident history actually looks like
Anthropic posts a lot of incidents. We counted 40 in the 31 days to 14 August 2026, roughly nine a week. That reads worse than it is once you look at how long they last: the median one closed in 67 minutes, three quarters were done inside two hours, and the longest ran eight.
Claude Code barely ever breaks by itself. All 40 named Claude Code as an affected component, and 37 of them named the Claude API alongside it. When the CLI stops answering, the thing underneath it is usually what went.
Most of the titles name a model rather than a service: "Elevated errors on Claude Sonnet 5", "Degraded performance for Claude Opus 5", "Elevated errors for Opus 5". That is why switching models is the first thing to try here and not a last resort. A model-scoped incident leaves the rest of the fleet answering.
40 incidents, 15 July – 14 August 2026. Counted from status.claude.com/api/v2/incidents.json, pulled . The live board at the
top of this page is current; these numbers are a snapshot and will drift.
Is Claude Code down, or is it just you?
- A 529 is Anthropic's "overloaded" response. During a listed incident it means capacity, not a mistake in your setup, and it usually clears on its own.
- A 401 or 403 is yours. The token expired or the session went stale — run /login again before you open a status page.
- If Claude Code hangs but claude.ai answers in the browser, rule out your own network path. A proxy or VPN that buffers responses will stall the streaming connection the CLI holds open, and a stall looks identical to an outage.
What to do while Claude Code is down
- Check whether the incident names one model. If so, switch models for the session (e.g. /model in Claude Code) and keep working.
- A 529 ("overloaded") or repeated 500s during a listed incident is the outage, not your setup. A 401/403 is your auth — re-run /login.
- If Claude Code is listed as degraded but the API is green, the CLI's own backend is the problem — the API and claude.ai will still work.
- For long outages, an editor-agnostic fallback like Aider pointed at a different provider keeps you shipping.
Frequently asked questions
How do I check if Claude Code is down?
Look at the live board above — it reads Anthropic's official status API (status.claude.com) directly in your browser, including the dedicated "Claude Code" component. If the board is unreachable, open status.claude.com itself.
How often is Claude Code down?
More often than you would guess, and for less time than you would fear. Anthropic logged 40 incidents in the 31 days to 14 August 2026 — about nine a week — but the median one was resolved in 67 minutes and three quarters closed inside two hours. Most were scoped to a single model rather than the whole service.
Why does Claude Code fail while claude.ai works?
Anthropic tracks them as separate components. Claude Code depends on the Claude API and its own backend, so a component-level incident can hit the CLI while chat on claude.ai stays green.
What do 529 errors in Claude Code mean?
A 529 is Anthropic's "overloaded" response. During a listed incident it simply means capacity is constrained — retrying later or switching to a less-loaded model usually gets you moving before the incident fully resolves.